google-agents-cli-observability

Observability Guide

Cloud Trace works out of the box — no infrastructure needed. Prompt-response logging and BigQuery Agent Analytics require Terraform-provisioned infrastructure (service account, GCS bucket, BigQuery dataset). Run agents-cli infra single-project --apply --project PROJECT_ID to provision these resources. Go projects get the BigQuery telemetry stack too; the GCS completion upload behind prompt-response logging and the BigQuery Agent Analytics plugin are Python only. See references/cloud-trace-and-logging.md for details, env vars, and verification commands. If your project isn't scaffolded yet, see /google-agents-cli-scaffold first.

Deployment order

Do you want Terraform-managed observability?	Action
No	Run agents-cli deploy directly (works for all targets).
Yes	Run agents-cli infra single-project --apply first, then agents-cli deploy.

If you already ran agents-cli deploy imperatively: Do not apply Terraform afterward. For Agent Runtime in Gemini platform, Terraform owns the whole engine (service account, deployment spec, env vars), so a later apply can't be reconciled without taking over the resource. Either delete the deployment and start over, or keep the SDK-deployed instance — skip infra single-project and set the observability env vars by re-running agents-cli deploy --update-env-vars "KEY=VALUE,..."; deploy matches the existing Reasoning Engine by display name and updates it in place, preserving env vars set outside the deploy. You must also grant its service account the telemetry IAM roles the Terraform module would otherwise provision: roles/storage.admin (write completions to the logs bucket), roles/logging.logWriter, roles/cloudtrace.agent, plus roles/bigquery.dataOwner + roles/bigquery.jobUser when scaffolded with --bq-analytics. The full set lives in deployment/terraform/single-project/iam.tf (from app_sa_roles) and telemetry.tf. Terraform-managed env vars aren't available in this mode.

Reference Files

File	Contents
references/cloud-trace-and-logging.md	Scaffolded project details — Terraform-provisioned resources, environment variables, verification commands, enabling/disabling locally
references/bigquery-agent-analytics.md	BQ Agent Analytics plugin — enabling, key features, GCS offloading, tool provenance
references/adk-docs.md	ADK: adk.dev pages to fetch for detail beyond this skill
references/feedback-mechanism.md	Adding a user-feedback endpoint — request model, structured logging, log sink → BigQuery

---

Observability Tiers

Choose the right level of observability based on your needs:

Tier	What It Does	Scope	Default State	Best For
Cloud Trace	Distributed tracing — execution flow, latency, errors via OpenTelemetry spans	All templates, all environments	Always enabled	Debugging latency, understanding agent execution flow
Prompt-Response Logging	GenAI interactions exported to GCS, BigQuery, and Cloud Logging	Scaffolded ADK Python and Claude projects	Disabled locally, enabled when deployed	Auditing LLM interactions, compliance
BigQuery Agent Analytics	Structured agent events (LLM calls, tool use, outcomes) to BigQuery	ADK Python agents with the plugin enabled; Claude template agents	Opt-in (--bq-analytics at scaffold time)	Conversational analytics, custom dashboards, LLM-as-judge evals
Third-Party Integrations	External observability platforms (AgentOps, Phoenix, MLflow, etc.)	Any OpenTelemetry-instrumented agent	Opt-in, per-provider setup	Team collaboration, specialized visualization, prompt management

Ask the user which tier(s) they need — they can be combined. Cloud Trace is always on; the others are additive.

---

Cloud Trace

Scaffolded agents use OpenTelemetry to emit distributed traces. Every agent invocation produces spans that track the full execution flow.

Span Hierarchy

ADK projects. These are ADK's span names; other frameworks emit their own (generate_content comes from the shared google-genai instrumentor either way).

invoke_workflow (top-level run)
  └── invoke_agent (one per agent in the chain)
        ├── call_llm (model request)
        │     └── generate_content (underlying GenAI model call)
        └── execute_tool (tool execution)

Setup by Deployment Type

Deployment	Setup
Agent Runtime	Automatic — exporters wired at startup, gated on GOOGLE_CLOUD_AGENT_ENGINE_ENABLE_TELEMETRY (set by deploy); exports to Cloud Trace/Logging + the Agent Runtime page in the Google Cloud console
Cloud Run / GKE (scaffolded)	Automatic — exporters wired at startup, exports to Cloud Trace/Logging
Cloud Run / GKE (manual)	Configure OpenTelemetry exporter in your app
Local dev	Works with agents-cli playground; traces visible in Cloud Console

Wired at app startup — ADK Python: get_fast_api_app(otel_to_cloud=True) in app/fast_api_app.py; ADK Go: setupObservability() in observability.go; other templates call their own. references/cloud-trace-and-logging.md has the details.

View traces: Cloud Console → Trace → Trace explorer

ADK: for detailed setup instructions (Agent Runtime CLI/SDK, Cloud Run, custom deployments), fetch https://adk.dev/integrations/cloud-trace/index.md.

---

Prompt-Response Logging

Captures GenAI interactions and exports to GCS (JSONL) and BigQuery (via log sinks + external tables). Content is governed by two independent tiers; the net Terraform-deploy default is full content in GCS/BigQuery, none in traces:

Tier	Captures	Controlled by	Default (Terraform deploy)
GCS/BigQuery completions	Full prompts/responses (the prompt-response logging feature)	OTEL_INSTRUMENTATION_GENAI_COMPLETION_HOOK=upload + LOGS_BUCKET_NAME	On — full content
Trace spans / Cloud Logging events	Span/event content	OTEL_INSTRUMENTATION_GENAI_CAPTURE_MESSAGE_CONTENT (plus ADK_CAPTURE_MESSAGE_CONTENT_IN_SPANS=false, ADK Python only)	Off — NO_CONTENT

The tiers are independent: GCS/BigQuery uploads capture full content whenever their upload vars are set and do not honor OTEL_INSTRUMENTATION_GENAI_CAPTURE_MESSAGE_CONTENT, which governs the traces/events tier only.

ADK Python reads it as the experimental-semconv enum:

• NO_CONTENT — no content in spans/events (scaffolded default)
• EVENT_ONLY — content in Cloud Logging events
• SPAN_ONLY / SPAN_AND_EVENT — content in trace spans
• true / false — invalid; fall back to NO_CONTENT

ADK Go reads the same variable as a boolean: "1" or "true" capture content, every other value — including the enum members above — elides it.

For the full mechanics (semconv opt-in, declarative Terraform config, env-var table, enabling/disabling, verification commands), see references/cloud-trace-and-logging.md. For ADK logging docs (log levels, configuration, debugging), fetch https://adk.dev/observability/logging/index.md.

---

BigQuery Agent Analytics Plugin

ADK projects. Optional ADK plugin that logs structured agent events to BigQuery. Enable with --bq-analytics at scaffold time. See references/bigquery-agent-analytics.md for details.

---

Third-Party Integrations

Many third-party observability platforms can ingest agent telemetry (via OpenTelemetry or custom instrumentation). The table below covers common ones; the full list is larger (see the pointer below it).

Platform	Key Differentiator	Setup Complexity	Self-Hosted Option
AgentOps	Session replays, 2-line setup, replaces native telemetry	Minimal	No (SaaS)
Arize AX	Commercial platform, production monitoring, evaluation dashboards	Low	No (SaaS)
Phoenix	Open-source, custom evaluators, experiment testing	Low	Yes
MLflow	OTel traces to MLflow Tracking Server, span tree visualization	Medium (needs SQL backend)	Yes
Monocle	1-call setup, VS Code Gantt chart visualizer	Minimal	Yes (local files)
Weave	W&B platform, team collaboration, timeline views	Low	No (SaaS)
Freeplay	Prompt management + evals + observability in one platform	Low	No (SaaS)

Ask the user which platform they prefer — present the trade-offs and let them choose. ADK: fetch a platform's setup page at https://adk.dev/integrations/<slug>/index.md (slugs for the table above: agentops, arize-ax, phoenix, mlflow-tracing, monocle, weave, freeplay); ADK has more observability integrations (Datadog, Galileo, LangWatch, Latitude, Future AGI, Respan, Zespan, …) — browse the complete, current list at https://adk.dev/integrations/ (observability topic). On other frameworks the OpenTelemetry-based platforms still work, but follow the platform's own setup docs.

---

Troubleshooting

Issue	Solution
No traces in Cloud Trace	Verify telemetry setup runs at startup and the SA has the cloudtrace.agent role. ADK Python: fast_api_app.py uses get_fast_api_app(otel_to_cloud=True); ADK Go: observability.go build the exporter manually. Agent Runtime additionally gates this on GOOGLE_CLOUD_AGENT_ENGINE_ENABLE_TELEMETRY.
Prompt-response data not appearing	Check LOGS_BUCKET_NAME is set; verify SA has storage.objectCreator on the bucket; check app logs for telemetry setup warnings
Content in traces/events (unwanted)	ADK Python: OTEL_INSTRUMENTATION_GENAI_CAPTURE_MESSAGE_CONTENT=NO_CONTENT keeps content out of spans/events. ADK Go: any value other than 1/true does, so unset it or set false. NOTE: GCS/BigQuery completions still capture full content — to stop that, remove LOGS_BUCKET_NAME/OTEL_INSTRUMENTATION_GENAI_COMPLETION_HOOK (drop the upload block in service.tf)
BigQuery Analytics not logging	ADK Python: verify the plugin is configured in app/agent.py; Claude: check the startup log for a BigQuery Agent Analytics warning. Both: check BQ_ANALYTICS_DATASET_ID env var is set
Third-party integration not capturing spans	Check provider-specific env vars (API keys, endpoints); some providers (AgentOps) replace native telemetry
Traces missing tool spans	ADK: tool execution spans appear under execute_tool (other frameworks use their own span names) — check trace explorer filters
High telemetry costs	Turn content capture off (NO_CONTENT in Python, false in Go); reduce BigQuery retention; disable unused tiers

---

Related Skills

• /google-agents-cli-deploy — Deployment targets, CI/CD pipelines, and production workflows
• /google-agents-cli-workflow — Development workflow, coding guidelines, and operational rules
• /google-agents-cli-adk-code — ADK API quick reference for writing agent code, Python and Go